Methodology & Data Sources

Our Mission

Statistics of the World provides free browsing and basic API access to economic, demographic, health, education, and environmental data for 217 countries. Each series retains its provider attribution and release context. Our added work is in harmonization: reconciling country names, ISO codes, and territorial definitions that differ between publishers, normalizing units and response formats, preserving observation-versus-projection labels when the source provides them, and writing analysis that puts the numbers in context. Coverage and update timing vary by provider and series; users should follow the original provider's citation and licensing requirements.

Primary Data Sources

Data Processing

Our ETL (Extract-Transform-Load) pipeline runs on a scheduled basis, fetching data from source APIs and storing it in our database. The pipeline:

• Extracts data from official APIs (IMF DataMapper, World Bank API, WHO GHO API, OECD SDMX API)
• Transforms data into a consistent format — standardizing country names, ISO codes, indicator IDs, and time periods
• Validates data for completeness and consistency, flagging anomalies for review
• Loads data into our PostgreSQL database, preserving full historical time series

We do not modify, adjust, or smooth source data. If the IMF reports GDP for Country X as $500 billion, that is exactly what we display. Any projections or estimates are clearly labeled as such and attributed to the original source.

Update Frequency

• IMF macroeconomic data: Updated within 48 hours of each WEO release (April, October)
• World Bank indicators: Updated weekly as new data becomes available
• Financial market data: Updated daily during trading hours
• WHO health data: Updated as new releases are published (typically annual)

Data Limitations

All data has limitations that users should understand:

• Lag: Most World Bank indicators are 1-2 years behind the current year. IMF data includes current-year estimates and multi-year projections.
• Coverage gaps: Not all indicators are available for all countries. Small states, conflict zones, and authoritarian regimes may have incomplete data.
• Revisions: GDP and other economic data are regularly revised as more complete information becomes available. Historical data may change.
• Comparability: Despite harmonization efforts, national statistical methodologies differ. Cross-country comparisons should account for these differences. Some measures follow each country's own definition: the IMF's unemployment rates (who counts as unemployed, which ages, survey or registry), consumer price inflation (each country's own price index), government debt and balance (which parts of government are covered) and poverty at national poverty lines. Ranking pages built on these say so above the table, and where a harmonized alternative exists (the World Bank's modelled ILO unemployment estimates, its international poverty line) they name it.
• Informal economy: GDP and employment statistics may undercount informal economic activity, particularly in developing countries.

Editorial Content

Long-form editorial content — country economy pages, ranking analyses, comparison write-ups, and blog articles — is written and reviewed in-house against the underlying series. Short descriptive summaries on data pages are generated directly from the underlying series and restate what the charts show; they make no claims beyond the data. Editorial content is reviewed for accuracy and updated when significant economic developments occur. We cite specific data sources throughout and link to the original data where available.

API Access

Our free API provides programmatic access to all data in JSON format. Basic usage requires no authentication (1,000 requests/day per IP). Free API keys add usage tracking and a dashboard; paid tiers raise the limits. Academic and research use is encouraged.

Recorded source-comparison pilot

Our source comparison workspace compares ten IMF and World Bank macro concepts for twenty economies in matching reference years. Differences are descriptive; they do not establish a cause, reporting misconduct or which source is correct. General-government and central-government debt have different scopes, so their difference is withheld. Unrecognized units remain raw until reviewed.

The archive preserves inserted, changed, withdrawn and missing stored rows, plus definition metadata, from its installation baseline onward. It records transaction-final SOTW state. Source publication dates, publisher vintages and database commit availability are not established; this archive alone cannot certify what was publicly knowable on a historical date. Earlier history is explicitly unavailable.

Evidence downloads pin the recording cutoff, archive revision and processing version, retain attribution, and include a data checksum. Archived metadata is distinguished from reference catalog fallbacks and their review dates. Revision timelines retain native stored values because earlier unit changes may prevent reliable normalization. Original response bodies are retained privately where captured; a hash identifies bytes and does not certify economic accuracy.

How fresh is SOTW against other sites?

The economic data freshness benchmark checks six free economic data sites, SOTW included, against the newest official release for 50 headline cells, on rules fixed in October 2026, with every page read listed. SOTW runs it and is one of the sites scored; the page says so and shows where SOTW loses.

Corrections & Feedback

If you identify a data error, outdated information, or have suggestions for improvement, please contact us. We take data accuracy seriously and will investigate all reports promptly.